Deep Learning Foundations & Azure Environments

This course is part of Microsoft Deep Learning Engineering with Azure Professional Certificate

Instructor: Microsoft

What you'll learn

Implement feedforward neural networks in PyTorch, including forward/backward passes, loss functions, and optimizer configuration
Build custom nn.Module classes and configure DataLoader pipelines with mixed precision training and torch.compile optimization
Configure Azure ML workspaces, compute clusters, and GPU targets using the Azure ML SDK v2
Submit training jobs, track experiments with MLflow, and register models using the Azure ML model registry

Details to know

Shareable certificate

Add to your LinkedIn profile

Assessments

13 assignments¹

AI Graded see disclaimer

Taught in English

See how employees at top companies are mastering in-demand skills

Learn more about Coursera for Business

logos of Petrobras, TATA, Danone, Capgemini, P&G and L'Oreal

Build your Machine Learning expertise

This course is part of the Microsoft Deep Learning Engineering with Azure Professional Certificate

When you enroll in this course, you'll also be enrolled in this Professional Certificate.

Learn new concepts from industry experts
Gain a foundational understanding of a subject or tool
Develop job-relevant skills with hands-on projects
Earn a shareable career certificate from Microsoft

There are 7 modules in this course

Production-grade deep learning starts with solid engineering foundations. This course builds the core PyTorch and Azure ML skills needed to build, train, and track deep learning models in enterprise environments.

You'll implement feedforward neural networks from scratch, including forward and backward passes, loss functions, and optimizers, then move into custom PyTorch module design, efficient DataLoader pipelines, and mixed precision training with torch.compile. You'll configure Azure ML workspaces, compute clusters, and GPU targets using the Azure ML SDK v2, manage training jobs, and track experiments with MLflow. By the end of this course, you'll be able to build custom PyTorch training pipelines, configure Azure ML environments, and monitor deep learning workflows from end to end. This course is designed for engineers with intermediate Python and hands-on ML experience, ready to build deep learning skills on Azure. Labs use a dual-path design: a gradable track with precomputed outputs and logs supports workflow mastery without live GPU provisioning, while an optional live path is available for enterprise learners with Azure subscriptions and GPU quotas.

Establish a robust understanding of core deep learning architecture. You will explore and manually code computational graphs, evaluating how data moves through forward and backward passes.

What's included

1 video3 readings1 assignment

Move beyond architecture setup and focus on convergence. You will configure and test advanced optimization and activation techniques to evaluate their direct impact on model performance.

What's included

2 videos2 readings3 assignments

2 videosTotal 16 minutes

Activation Functions and Gradient Flow7 minutes
Loss Functions and Loss Landscapes9 minutes

2 readingsTotal 15 minutes

Gradient Descent and Advanced Optimizers10 minutes
Configuring Optimizers and Profiling in PyTorch5 minutes

3 assignmentsTotal 75 minutes

Graded Quiz: Neural Network Foundations30 minutes
Hands-on Activity: Tabular Classification Network Training Report30 minutes
Knowledge Check: Loss and Optimizers15 minutes

Move beyond pre-packaged solutions to establish modular control over your network components. You will structure custom nn.Module classes and explicitly interact with PyTorch's autograd engine.

What's included

2 videos3 readings1 assignment

2 videosTotal 11 minutes

Why Custom Modules Drive Deep Learning Scale3 minutes
Inside PyTorch Autograd and Graphs8 minutes

3 readingsTotal 31 minutes

Designing Configurable nn.Module Classes5 minutes
Building Custom nn.Module Classes6 minutes
Hands-on Activity: Constructing a Custom Neural Architecture20 minutes

1 assignmentTotal 15 minutes

Knowledge Check: Custom Modules & Autograd15 minutes

Tackle performance bottlenecks at the source. You will build data ingestion pipelines and implement PyTorch's native acceleration tools to train networks significantly faster.

What's included

1 video2 readings3 assignments

1 videoTotal 10 minutes

Datasets, Mixed Precision, and torch.compile10 minutes

2 readingsTotal 9 minutes

The Data Loading Bottleneck3 minutes
Configuring Mixed Precision and Speed Optimization6 minutes

3 assignmentsTotal 65 minutes

Graded Quiz: PyTorch Deep Dive30 minutes
Hands-on Activity: Image Classification Performance Comparison20 minutes
Knowledge Check: DataLoaders & Optimization15 minutes

Establish your deep learning infrastructure in the cloud. You will learn to programmatically define and provision Azure ML workspaces, map scalable modern NCasT4-series GPU compute targets and secure your execution environments using production-ready container registries.

What's included

1 video3 readings1 assignment

Execute and track your deep learning training pipelines. You will format PyTorch training scripts as Azure ML command jobs, connect MLflow for seamless metric tracking, and securely map data assets.

What's included

1 video1 reading3 assignments

Synthesize your foundational engineering skills to provision a cloud environment and train a baseline neural network. You will design an end-to-end blueprint that integrates custom PyTorch modules, optimized DataLoaders, and Azure ML scaling configurations.